activity
20242026
collaborators

10 papers

cs.CV2026

GeoT2V-Bench: Benchmarking 3D Consistency in Text-to-Video Models via 3D Reconstruction

Chenrui Fan, Paolo Favaro

Camera-prompted text-to-video (T2V) models are increasingly used to synthesize virtual camera captures, such as orbiting objects or moving through static scenes. For these outputs,…

cs.CV2026

World Model Self-Distillation: Training World Models to Solve General Tasks

Sebastian Stapf, Pablo Acuaviva Huertos, Aram Davtyan +1

Pretrained video generators are promising visual world models that exhibit emergent task-solving abilities; however, their reliance on detailed textual descriptions limits their di…

cs.LG2026

Faster Inference of Flow-Based Generative Models via Improved Data-Noise Coupling

Aram Davtyan, Leello Tadesse Dadi, Volkan Cevher +1

Conditional Flow Matching (CFM), a simulation-free method for training continuous normalizing flows, provides an efficient alternative to diffusion models for key tasks like image…

cs.CV2026

Communication-Inspired Tokenization for Structured Image Representations

Aram Davtyan, Yusuf Sahin, Yasaman Haghighi +4

Discrete image tokenizers have emerged as a key component of modern vision and multimodal systems, providing a sequential interface for transformer-based architectures. However, mo…

cs.CV2025

Rethinking Visual Intelligence: Insights from Video Pretraining

Pablo Acuaviva, Aram Davtyan, Mariam Hassan +4

Large language models (LLMs) have demonstrated that large-scale pretraining enables systems to adapt rapidly to new problems with little supervision in the language domain. This su…

cs.LG2025

KOALA++: Efficient Kalman-Based Optimization with Gradient-Covariance Products

Zixuan Xia, Aram Davtyan, Paolo Favaro

We propose KOALA++, a scalable Kalman-based optimization algorithm that explicitly models structured gradient uncertainty in neural network training. Unlike second-order methods, w…